System Design — Suite anchor

Systema

Make complex problems thinkable. Personas that won't lie to you about their own confidence.

Coming Q3 2026 to the Mac App Store

What it does

Systema is the labsly Suite's System Design app. Map stakeholders, surface hidden structures with the iceberg model, trace causal dynamics with reinforcing/balancing loop diagrams, scan for interventions with STEEP, and build a testable theory of change. Talk to your stakeholders via Persona Simulator — every output labeled with confidence, source, and evidence so you know what the AI doesn't know.

Then run it. Any loop diagram can be promoted to a working model with real delays, drawable curves, and an integrator that verifies its own numerics rather than asking you to trust them. Runs come back as a band of sampled futures instead of one confident line. Write guardrails and they are checked across every future, not just the headline run. Drop in a CSV and the model has to answer to your data. Ask what drove any moment and it shows the arithmetic.

Modules

Stakeholder Map

Shipped

Four-layer power-distance model with progressive disclosure.

Iceberg Model

Shipped

Events / Patterns / Structures / Mental Models depth analysis.

Causal Loop Diagrams

Shipped

With reinforcing/balancing loop detection and 8-archetype matching.

STEEP Boards

Shipped

Social, Technological, Economic, Environmental, Political scanning.

Theory of Change

Shipped

Five-column logic model with assumption tracking.

Simulation engine

Shipped

Promote a loop diagram to a running model. RK4 integration that cross-checks itself at every half-step, real delays and smoothing, and drawable lookup curves.

Ensemble bands

Shipped

Every run is a fan of sampled futures rather than one confident line, with sensitivity ranked from the band's own draws.

Guardrails

Shipped

Assertions checked on every run across every sampled future — model CI, with CLI and MCP surfaces so agent-built models arrive pre-tested.

Ground Truth

Shipped

Drop a CSV on the band and read how much of reality actually falls inside your claim.

Why-trace

Shipped

Ask what drove any moment and get exact flow arithmetic plus ranked causes, live under the scrubber.

Field Guide

Shipped

Eight test-verified archetypes you can open as a live, editable system instead of a picture in a book.

Challenges

Shipped

Goal-framed play with band-aware ROBUST / FRAGILE / NOT-YET verdicts.

MCP Server

Built — free

27 tools exposed to AI coding agents, including simulate, critique, and assertion tools. Built; published free at launch.

Persona Simulator

Q3 2026 — Flagship

Talk to your stakeholders. Confidence + provenance on every response.

CLD Critique Assistant

Q3 2026

LLM reviews your loops for archetype match and leverage point opportunities.

Domain Sense

Q3 2026

Clear / complicated / complex / chaotic classifier with disposition logic.

Live Critique Companion

Q4 2026

Real-time Socratic watcher on your model.

Who it's for

Independent strategists, researchers, senior PMs, design directors, indie consultants, foundation program officers, think-tank fellows, PhD students. People who already think systemically and want a tool that respects their craft.

How it's different

vs Vensim / Stella: AI-native from the ground up, Mac-native craft, and a simulation core that checks its own numerics instead of asking you to trust them.
vs Kumu: Kumu refuses AI. We embrace it — with epistemic honesty.
vs Miro / FigJam: Structured by default, not a sticky-note pile.
vs ChatGPT alone: Persistent structured projects; AI reasons over your model, not over guesses.

Availability

Systema is in active development for a Q3 2026 Mac App Store release. The MCP server is built and will be free for agents when it is published — it is not on PyPI yet. Pricing will be published before the App Store listing opens. MCP server details.

Get launch updates

Request launch updates

Direct, occasional updates from the person building the apps.

Methodology acknowledgments

Systema's analytical modules build on the work of Donella Meadows (iceberg model, leverage points), Jay Forrester (system dynamics, causal loops), Peter Senge & Daniel Kim (system archetypes), Garrett Hardin (Tragedy of the Commons), Urie Bronfenbrenner (ecological stakeholder model), and the broader logic-model tradition (Kellogg Foundation, Aspen Institute, ActKnowledge). None of these originators is affiliated with or endorses Systema or labsly.